Walking dynamics are symmetric (enough).

Area of Science:

• Biomechanics
• Dynamical Systems Theory
• Human Locomotion Analysis

Background:

• Rhythmic biological phenomena, including locomotion, are often modeled as rhythmic dynamical systems.
• Simplifications in dynamical systems modeling, such as assuming bilateral symmetry in human locomotion, are common.
• The validity of assuming symmetry in human walking dynamics around the steady state (limit-cycle) remains to be rigorously tested.

Purpose of the Study:

• To statistically examine human walking dynamics for significant asymmetries around the limit-cycle.
• To evaluate the impact of assuming bilateral symmetry on the predictive accuracy of walking models, even in the presence of asymmetries.

Main Methods:

• Adaptation of statistical cross-validation techniques to analyze walking dynamics.
• Comparison of model performance with and without the assumption of bilateral symmetry.
• Focus on steady-state (limit-cycle) walking dynamics.

Main Results:

• Statistically significant asymmetries were identified in the dynamics of human walking.
• Models that ignored these identified asymmetries demonstrated improved consistency and predictive power.
• The simplification of neglecting evident system characteristics paradoxically led to a superior model.

Conclusions:

• Bilateral symmetry is not strictly present in human walking dynamics.
• Ignoring identified asymmetries can paradoxically enhance the robustness and predictive capability of dynamical models.
• Model simplification by neglecting certain characteristics can be a powerful strategy for improving biological system modeling.